Rent Seeking and the Unveiling Of'de Facto'institutions: Development and Colonial Heritage Within Brazil
Total Page:16
File Type:pdf, Size:1020Kb
NBER WORKING PAPER SERIES RENT SEEKING AND THE UNVEILING OF 'DE FACTO' INSTITUTIONS: DEVELOPMENT AND COLONIAL HERITAGE WITHIN BRAZIL Joana Naritomi Rodrigo R. Soares Juliano J. Assunção Working Paper 13545 http://www.nber.org/papers/w13545 NATIONAL BUREAU OF ECONOMIC RESEARCH 1050 Massachusetts Avenue Cambridge, MA 02138 October 2007 The authors gratefully thank Flávia Chein Feres for essential help in managing and sharing the basic geo-referenced dataset on Brazilian municipalities, Cláudio Egler for providing the geo-referenced data on the coastline and the projection of the Brazilian map in kilometers, and Jesus Fernando Mansilla Baca for providing the data on temperatures and types of soil. The paper also benefited from comments and suggestions from Marcelo de Paiva Abreu, Roger Betancourt, Filipe Campante, Cláudio Ferraz, Karla Hoff, Peter Murrell, Edson Severnini, Dietrich Vollrath, and seminar participants at Columbia University, IMF, IPEA-Rio, PUC-Rio, Johns Hopkins University/Center for Global Development, University of Delaware, University of Houston, University of Maryland-College Park, University of Pittsburgh, World Bank, the 2007 Panel of the LACEA Political Economy Group (Cartagena), the 2007 NBER Summer Institute on Income Distribution and Macroeconomics (Cambridge), and the 2007 LACEA Annual Meeting (Bogotá). The views expressed herein are those of the authors and do not necessarily represent the view of the institutions to which they are affiliated, or those of the National Bureau of Economic Research. © 2007 by Joana Naritomi, Rodrigo R. Soares, and Juliano J. Assunção. All rights reserved. Short sections of text, not to exceed two paragraphs, may be quoted without explicit permission provided that full credit, including © notice, is given to the source. Rent Seeking and the Unveiling of 'De Facto' Institutions: Development and Colonial Heritage within Brazil Joana Naritomi, Rodrigo R. Soares, and Juliano J. Assunção NBER Working Paper No. 13545 October 2007 JEL No. N26,O17,O40,P14,P28 ABSTRACT This paper analyzes the roots of variation in de facto institutions, within a constant de jure institutional setting. We explore the role of rent-seeking episodes in colonial Brazil as determinants of the quality of current local institutions, and argue that this variation reveals a de facto dimension of institutional quality. We show that municipalities with origins tracing back to the sugar-cane colonial cycle -- characterized by a polarized and oligarchic socioeconomic structure -- display today more inequality in the distribution of land. Municipalities with origins tracing back to the gold colonial cycle -- characterized by an over-bureaucratic and heavily intervening presence of the Portuguese state -- display today worse governance practices and less access to justice. The colonial rent-seeking episodes are also correlated with lower provision of public goods and lower income per capita today, and the latter correlation seems to work partly through worse institutional quality at the local level. Joana Naritomi Juliano J. Assunção The World Bank Departamento de Economia 1818 H Street, NW Catholic University of Rio de Janeiro (PUC-Rio) Washington, DC 20433 USA Rua Marquês de São Vicente, 225 [email protected] 22451-900 Rio de Janeiro, RJ Brazil Rodrigo R. Soares [email protected] Departamento de Economia Catholic University of Rio de Janeiro (PUC-Rio) Rua Marquês de São Vicente, 225 22451-900 Rio de Janeiro, RJ Brazil and NBER, and the University of Maryland [email protected] 1. Introduction This paper analyzes the causes of variation in local institutions, within a constant ‘macro- institutional’ setting. We show that variation in institutional quality across Brazilian municipalities is partly inherited from the distinct colonial histories experienced by different areas of the country.1 Specifically, we explore the role of rent-seeking colonial episodes in determining the quality of current local institutions along four dimensions: distribution of economic power through distribution of endowments, persistence of political power, quality of local government practices, and access to the justice system. Our results show that municipalities with origins tracing back to the sugar-cane colonial cycle – characterized by a polarized and oligarchic socioeconomic structure – display today more inequality in the distribution of land. Municipalities with origins connected to the gold colonial cycle – characterized by an over- bureaucratic and heavily intervening presence of the Portuguese state – display today worse governance practices and less access to justice. Areas associated with the colonial rent-seeking episodes also experience lower provision of various types of public goods. In addition, exposure to the colonial rent-seeking episodes is correlated with lower income per capita today, and this correlation seems to work partly through worse governance practices and less access to justice at the local level. The role of institutions as key determinants of development has received increased attention in recent years. After the initial work of North (1991) and Engerman and Sokoloff (1997), a vast array of cross-country empirical literature developed following the footsteps of the seminal contributions of Acemoglu, Johnson, and Robinson (2001 and 2002).2 Much of this literature has evolved around the idea that the geographic pattern of development observed across countries – summarized in the relationship between distance to the equator and income per capita reproduced here in Figure 1 – reflects different institutional arrangements, inherited from different experiences of colonization. According to the emerging consensus in this literature, geographic conditions were associated with particular paths of colonization, which in turn translated into the establishment of different types of institutions. Institutions then, through 1 Municipalities are the smallest political and administrative units in Brazil. They have administrative autonomy, are governed by a mayor and a chamber of representatives, and are something in between US counties and cities. Municipalities raise certain taxes, receive transfers from the federal government, and decide on part of the expenditures along several different dimensions (education, health, infrastructure, etc.). 2 See, for example, Easterly and Levine (2003), Rodrik, Subramanian, and Trebbi (2004), and Acemoglu and Johnson (2005). Other literature contends that the geographic pattern of development reflects indeed the direct impact of geography on income per capita, through its effects on the disease environment, agricultural productivity, and access to trade (see, for example, Gallup, Sachs, and Mellinger, 1999). 1 their effects on property rights, political competitiveness, and governance, led to good policies and, ultimately, development. In this view, the adoption of distinct ‘macro-institutions’ – determined at the country level, and related to the political and judicial systems and to the enforcement of laws – would be the intervening force in the observed correlation between geography and development (see, for example, Acemoglu, Johnson, and Robinson, 2005 and Persson and Tabellini, 2004). The geographic pattern of development within Brazil raises a series of questions in relation to the interpretation of the cross-country evidence and the conclusions of this literature. Figure 2 replicates the typical scatter-plot of distance to the equator and income per capita for the case of Brazilian municipalities. As in the cross-country context, municipalities closer to the equator have systematically lower income per capita. In fact, the relationship between latitude and income is stronger and tighter within Brazil than across countries: the R2 of this relation across Brazilian municipalities is 0.56 and the coefficient on latitude is 0.053, while across countries the R2 is 0.32 and the coefficient is 0.038 (cross-country data from the Penn World Tables; municipality data presented in section 3). As will be seen later on, the geographic pattern of development within Brazil is even more striking than this simple scatter-plot reveals. A more complete set of geographic variables explains up to 65% of the variation in income per capita across municipalities (Table 1, in section 3, presents this result). At the same time, Brazil is a country that shares a single colonizer and a single language, and has a very centralized federal system. The ‘macro-institutions’ typically highlighted in the interpretation of the cross-country evidence, as well as the historical variables identified as their sources of variation, are, and for the most part have always been, constant within the territory (constraints on executive, legal system, competitiveness of political system, colonizing power, legal tradition, etc).3 Therefore, the pattern observed within Brazil challenges the understanding that the correlation between geography and development reflects mostly the effect of climate and endowments on the type of ‘macro-institutions’ that were ultimately adopted at the country level. Two non-mutually exclusive possibilities arise from this challenge: geographic factors may be important on their own as direct determinants of long-term development, which would contradict the main consensus of the institutional literature; or, even within a constant de jure 3 The Brazilian territory has remained roughly the same since the 18th century. And, apart from brief and localized incursions of other